The Agent Skills Directory

[COMMAND_EXECUTION]: The shell scripts generate_diff_gifs.sh and sdg_tolerance_diff_gifs.sh execute system utilities including composite, convert (ImageMagick), and jq to perform image differencing and parse validation metrics.
[DYNAMIC_EXECUTION]: The script scripts/sdg_tolerance_diff_gifs.sh generates and executes an embedded Python script using a heredoc. This script is used at runtime to process numeric depth maps (.npy files) using the NumPy library.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data files which constitutes a potential surface for indirect prompt injection.
Ingestion points: The skill reads image files (PNG), numeric data files (NPY), and JSON QA reports from directories specified by the user or agent during runtime.
Boundary markers: No specific boundary markers or instructions to ignore embedded content are present for the processed data files.
Capability inventory: The skill is granted shell and filesystem permissions, allowing it to execute local commands and perform file operations (reading validation data, writing output GIFs, and cleaning up temporary files).
Sanitization: The skill relies on standard parsing logic of the underlying tools (ImageMagick, jq, numpy) and does not perform additional content validation or sanitization of the ingested data.

validation-diff-gifs